Out of stock in peak season: the four exits and what each one costs

Almost everything written about stockouts is about avoiding them. Fair enough, that is the right emphasis, and most of the work lives there. But there is a moment no post covers and every seller lives through at least once: it is December 11, your best SKU has 40 units, it sells 35 a day, your restock lands January 7, and the question is not how to avoid it. It already happened. The question is what you do tomorrow.

That moment almost always gets decided badly, for an understandable reason: it gets decided in a rush and with no written criteria. The seller flips into reaction mode, drops the price to “move what’s left” — which is exactly backwards — leaves campaigns running because nobody remembered to switch them off, and when stock hits zero the listing pauses itself and the ranking history that took months to build is gone.

The good news is that the exits are countable. There are four, each has a different cost, and you can decide which one applies with two or three numbers. What you need is the criteria written down before the emergency, because in peak season there is no time to think it through.

This post is about those four exits, how to choose between them, and the two things to do no matter which one you pick.

first: stop calling it a stockout and start counting days

Before picking an exit you need one number, and it is not “how many units do I have”. It is how many selling days are left at today’s velocity, and how many days until the restock lands. The gap between those two numbers is the hole, and the size of the hole decides everything.

With the example above: 40 units at 35 a day is a little over one day of coverage. The restock lands in 27 days. The hole is 26 days in the middle of peak season. That is not an inventory problem: it is the month’s loss, and it should be treated as such.

But notice the detail that changes the decision: today’s sales velocity is not last month’s. A SKU that sold 8 a day in October and sells 35 in December does not have “five days of inventory” by the quarter’s average: it has one. Computing coverage with a stale velocity is the number-one reason peak-season stockouts surprise people who were actually watching their numbers.

And there is a second number almost nobody calculates that belongs right next to it: how much net margin switches off when that SKU stops selling. A stockout on a thin-margin product and one on the product carrying the month’s profit do not deserve the same urgency or the same spend to fix, even though a flat SKU list makes them look identical.

Glossary: stockout, when the listing has nothing left to sell →

exit 1: pause and protect

Pausing is the exit that sounds like defeat and is often the right call. It means switching the listing off — or letting it hit zero, which has the same effect on MercadoLibre, where the system pauses it automatically — and stopping all spend on it.

What it costs you. The sales of the hole days, which never come back. And exposure: a paused listing keeps its history of visits, questions and prior sales, but stops appearing in search. While it is out, the category’s traffic goes to other sellers, and when you return you compete against listings that never broke stride.

What it saves you. Cancellations. And that is far more than it looks, because a sale you cancel for lack of a unit is not a lost sale: it is a negative reputation event that weighs inside the rolling window measuring your performance. On MercadoLibre that window is 60 days and seller-cancelled sales are one of the three thermometer factors. On Amazon, pre-fulfillment cancellations and order defects land in account health. In both cases: selling what you do not have is worse than not selling.

When it is the right exit. When the hole is long, when you have no honest way to fulfil, and when the SKU is not the one carrying your month. Also when cancellation risk is high because your inventory is not well synced across channels.

What to do when pausing that almost nobody does: switch off that SKU’s campaigns the same day. Leaving ads running against a paused or out-of-stock listing means paying for clicks that cannot convert. It is the easiest waste to avoid all season and the one most often forgotten.

exit 2: raise the price and ration

This is the counterintuitive exit and the one most frequently optimal. If you have 40 units left and the season has 26 more days to run, the price you set is wrong: it was calibrated for a demand that now vastly exceeds your supply.

Raising the price does three things at once. It slows sales velocity, which stretches the coverage of those 40 units. It raises the per-unit margin on the ones you do sell. And it keeps the listing active, so you lose neither history nor position.

What it costs you. Conversion, relative visibility against cheaper competitors, and on Amazon the risk of losing the Buy Box if you are competing for it on the same ASIN. That last point is the one to check first: if raising the price pushes you out of the buy box, you did not stretch coverage, you switched the sales off entirely.

What it saves you. Cancellations and the pause. And it leaves you holding a decision instead of a fait accompli.

When it is the right exit. When the hole is short or medium, when the SKU is not in a price war where one peso up takes you out of the game, and when the product has enough differentiation to hold a higher price. In December that condition holds more often than in any other month: the gift buyer has less price elasticity and more urgency.

The ordered version of this move, with steps and dates instead of one jump, is covered in laddered offers and in the automatic price calendar.

exit 3: switch fulfillment channels

If the inventory that ran out is the FBA or Full pool, but you still have units in your own warehouse or at your 3PL, there is one more question before pausing: can you keep selling that listing with your own shipping?

This is the exit that rescues the most value when it can be executed, because you lose neither the listing, nor the position, nor cancel anything. But it carries hard conditions.

What it costs you. Cost per shipment goes up, often substantially. Your handling time enters the promise calculation, and peak season is exactly when it degrades. And the estimated delivery date the buyer sees stretches, which cuts conversion.

What to verify first. That the SKU’s net margin can absorb the pricier freight. That your packing capacity can handle the extra volume without blowing up your handling time. And that your delivery promise stays keepable: switching to self-fulfillment and then missing the date is worse than having paused, because you add late shipments to what was only an inventory problem.

When it is the right exit. When you have real stock outside the marketplace network, when the SKU’s margin is wide, and when the hole falls in the first half of the month, while transit times are still reasonable. In the last week before Christmas this exit almost always stops being viable, precisely because of the delivery promise.

exit 4: take the hit, with your eyes open

The fourth exit is to keep selling knowing you will fail to fulfil some orders. It exists and sometimes it is chosen: when the SKU defines your month, when the restock is in transit with a confirmed date, and when the hole is a few days.

What it costs you. Cancellations or late shipments, claims, negative ratings, and all of it charged to a window that drags for weeks. On MercadoLibre, inside the 60 days that measure your reputation. On Amazon, inside the account health indicators. If your thermometer or your account health is already bruised, this exit is not available: the risk is not a bad rating, it is the account.

When it makes sense. Only with three conditions together: short hole, a confirmed and not estimated restock date, and enough reputation headroom to absorb the hit. If one of the three is missing, there is a better exit.

How to execute it without it being reckless. If you decide to absorb it, do not do it silently. Adjust the published handling time so the estimated date reflects reality. Answer buyer questions with the real date. A forewarned buyer who accepts the wait does not open a claim; a surprised one does.

what to do regardless of the exit

Two moves apply to all four, and they are the ones that rescue the most money.

Switch off that SKU’s advertising the same day. Whether the listing is paused, out of stock, repriced upward or carrying a longer promise, the ad spend that was calibrated for the previous scenario no longer makes sense. On a paused listing it is pure waste. In the other three it has to be recalibrated, not left on autopilot.

Move the demand to what you do have. A stockout on one SKU is an opportunity to push traffic to its variant, its sibling size or the equivalent product in your catalog. That requires your catalog to be structured as a tree rather than a flat list: if you do not know which SKU substitutes for which, you cannot make that move in time. And it requires seeing availability by family, because one sold-out size out of twelve is not the same as the size carrying half the sales.

Glossary: real net margin, with everything deducted →

the decision tree in four questions

To avoid deciding while on fire, four questions in order. The first clear “yes” gives you the exit.

1. Do I have stock of that SKU outside the marketplace network, and can the margin absorb my own freight? If yes, exit 3: switch fulfillment channels and keep the listing alive.
2. Is the hole a few days, with a confirmed restock date, and does my reputation have headroom? If yes, exit 4: absorb it, but adjust the published promise and answer with the real date.
3. Does the product have enough differentiation to hold a higher price without losing the Buy Box? If yes, exit 2: raise the price and ration.
4. If none of the above, exit 1: pause, kill the campaigns and protect the reputation.

All four questions are answered with three data points: real stock by location, net margin per SKU, and days of coverage at current velocity. With those three at hand, the tree takes two minutes to walk. Without them, every stockout becomes an argument.

what happens next: a stockout has a tail

The closing mistake is believing the problem ends when the restock lands. It does not, for two reasons.

The first is recovering position. Reappearing is not the same as selling the same again: a listing that was out has to rebuild momentum, and in December that happens against competitors who accumulated history while you were paused. It is worth planning the return — with price, with advertising and with enough stock not to break again — instead of simply reactivating.

The second is the reputation window. If you picked exit 4 or cancelled orders, the cost stays on the record until the rolling window releases it. Knowing that window’s end date changes decisions: there are moves worth postponing while the number still carries December’s hit.

And the third, which is the one that really matters: every stockout leaves data. How much velocity accelerated before it broke, how long the restock actually took, how much estimated sales were lost. That record is the input that puts next year’s reorder point in the right place. A stockout that goes undocumented repeats.

how this looks in iqseller

The panel does not decide for you and does not pause listings: that happens in Seller Central and in the MercadoLibre panel. What it solves are the three data points the decision tree runs on, which is exactly what tends to be missing when the emergency hits.

The Inventory module consolidates the stock that is yours wherever it sits — FBA, Full, your warehouse, your 3PL — with real availability by location, aligned with what each channel publishes. That is the answer to the first question in the tree, and without one view it is an exercise in phone calls and memory.

Forecast takes the sales velocity and the lead time you load and projects when each SKU hits zero, per channel. That turns the stockout from an event into a date, and moves the decision from December to November. Alerts flags when a SKU crosses its critical threshold or when velocity accelerates enough to pull that date forward, which in peak season is what separates deciding from reacting.

Profitability gives net margin per SKU with COGS you loaded, commissions from the Amazon settlement and MercadoLibre orders, FBA and Full fees, shipping, advertising, VAT and withholdings deducted. That is what decides whether the SKU can absorb your own freight and who gets your scarce attention. And the Parent → Model → SKU tree in the Catalog module is what lets you read the stockout by family and know which variant to move demand to.

What the panel does not do is report lost sales as hard data: that is an estimate built by projecting prior velocity over the days without stock, and it should always be presented as an estimate, never as a fact.

the minimum dashboard for the season

Five numbers, reviewed every other day from Buen Fin — November 13 to 17, 2026, according to the dates released by Mexico’s Ministry of Economy — through Three Kings’ Day:

• Days of coverage per SKU and per channel, computed with the last seven days’ velocity, not the quarter’s average.
• Seven-day velocity against thirty-day velocity, as a ratio. Any SKU above 2 is accelerating and its stockout date has already moved.
• Net margin per SKU, to rank who gets attention first when two SKUs break on the same day.
• Ad spend on out-of-stock or paused SKUs, which should be zero and almost never is.
• Cancellations and claims for the period, with the end date of the window that measures them.

What this changes is not the inventory you have: it changes the quality of the decision you make with the inventory you have left. And in peak season that is worth more than an extra shipment.